Cognitive Impairment in Schizophrenia Richard S.E. Keefe and Philip D. Harvey Contents 1 Cognitive Impairment in Schizophrenia and Its Clinical Relevance ....................... 12 1.1 Cognition in the Diagnosis of Schizophrenia ......................................... 12 1.2 Cognitive Deficits Are Found in Almost All Patients with Schizophrenia .......... 12 1.3 Cognitive Impairment Is Not Caused by Psychotic Symptoms ...................... 13 1.4 Cognitive Impairment Is an Important Cause of Functional Disability and Related Outcomes in Schizophrenia .............................................. 14 1.5 The Profile of Cognitive Impairment in Schizophrenia .............................. 15 1.6 Cognitive Impairment Precedes the Onset of Psychosis ............................. 18 1.7 Assessment of Cognition in Schizophrenia Treatment Studies ...................... 20 1.8 Early Phase Trials ...................................................................... 21 1.9 Functional Capacity .................................................................... 24 2 Treatments for Cognitive Impairment in Schizophrenia ................................... 25 2.1 Antipsychotic Effects on Cognition ................................................... 26 2.2 Pharmacological Augmentation as a Cognitive Enhancement Strategy ............. 27 2.3 Results of Cognitive Enhancement Efforts to Date .................................. 29 2.4 Cognitive Remediation as a Platform for Pharmacologic Studies ................... 29 3 Conclusions .................................................................................. 30 References ........................................................................................ 31 Abstract Cognitive functioning is moderately to severely impaired in patients with schizophrenia. This impairment is the prime driver of the significant disabilities in occupational, social, and economic functioning in patients with schizophrenia and an important treatment target. The profile of deficits in schizophrenia includes many of the most important aspects of human cognition: attention, memory, reasoning, R.S.E. Keefe (*) Duke University Medical Center, Durham, NC 27710, USA e-mail: [email protected]P.D. Harvey University of Miami Miller School of Medicine, Miami, FL 33136, USA e-mail: [email protected]M.A. Geyer and G. Gross (eds.), Novel Antischizophrenia Treatments, Handbook of Experimental Pharmacology 213, DOI 10.1007/978-3-642-25758-2_2, # Springer-Verlag Berlin Heidelberg 2012 11
28
Embed
Cognitive Impairment in Schizophrenia - Semantic Scholar · 1 Cognitive Impairment in Schizophrenia and Its Clinical Relevance ... symptom rating scales, ... (Green 1996) as well
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Theory-of-mind skills and social and emotion perception and recognition have been
the general focus of the literature on social cognition in schizophrenia. Theory of
mind is the ability to infer another’s intentions and/or to represent the mental states of
others. Individuals with schizophrenia perform poorly onmeasures of theory-of-mind
abilities (Tan et al. 2005; Corcoran et al. 1995; Drury et al. 1998). Facial affect
recognition and social cue perception are the two general areas into which studies of
social perception in schizophrenia can be broken down. Reviews of the literature on
facial affect recognition (Sarfati et al. 1997; Morrison et al. 1988; Penn et al. 1997)
suggest that individuals with schizophrenia have stable deficits on tests of facial
affect perception and that perception of negative emotions and fear may be particu-
larly impaired (Addington and Addington 2000; Penn et al. 1997; Pinkham et al.
2011; Edwards et al. 2001). Tests of social cue perception use more dynamic stimuli
that require multiple sensory modalities, such as watching people interacting. Patients
with schizophrenia show consistent impairments on these tasks (Gaebel and W€olwer1992; Bell et al. 1997). Social cognition is related to social impairments in schizo-
phrenia, even after controlling for performance on neurocognitive tasks (Corrigan
et al. 1990; Trumbetta and Mueser 2001). Path models have suggested that the
relations between social cognition and functional outcomes are complex, but that
social cognition may explain more of the direct variance in social functioning than
other aspect of cognitive performance (Penn et al. 1996).
1.6 Cognitive Impairment Precedes the Onset of Psychosis
Various methods for assessing the relationship between premorbid cognitive
impairment and later psychotic disorders have suggested that young people destined
18 R.S.E. Keefe and P.D. Harvey
to develop schizophrenia are modestly impaired on cognitive measures. However,
these deficits tend to be quite mild (Brekke et al. 2007) and their ability to help
predict psychotic disorders is under question. In a special circumstance, the longitu-
dinal follow-up of individuals who manifest prodromal symptoms (Reichenberg
et al. 2010), deficits on standard neuropsychological tests that are present at the time
of the development of the prodrome discriminate cases who go on to develop
psychosis from those who do not. However, impairment on these measures was
not able to contribute to the prediction of psychosis beyond clinical measures
implemented in the study.
Early work completed in the U.K. (Seidman et al. 2010) and Sweden (Jones et al.
1994) suggested that children who went on to develop schizophrenia as adults
differed significantly from the general population in a wide range of cognitive and
behavioral domains. Similar findings were generated from a population-based study
that investigated the risk of schizophrenia in the United States. Scores from grades
4, 8, and 11 on the Iowa Tests for 70 children who later developed schizophrenia
suggested that those children who later developed schizophrenia, test scores
dropped significantly between grades 8 and 11, corresponding with the onset of
puberty (David et al. 1997).
In Israel, a study of all adolescents between the ages of 16 and 17 years
suggested that cognitive functions are significantly impaired in those adolescents
who are later hospitalized for schizophrenia. These deficits thus precede the onset
of psychosis in young people destined to develop schizophrenia, and, along with
social isolation and organizational ability, cognitive deficits are a significant pre-
dictor of which young people will eventually develop a psychotic disorder (Fuller
et al. 2002). However, the mean level of performance of this group, at about the
35th percentile of the overall population, does not allow for a very strong predictive
signal on a case-by-case basis. In the young people who later experienced a first
episode of schizophrenia, their cognitive results in the prodrome suggested that
most of the cognitive impairment of schizophrenia occurs prior to the first psychotic
episode (Davidson et al. 1999).
Recent work from the Dunedin study in New Zealand, which tracked the
cognitive and mental health of a large group of individuals in a single geographical
location, suggests that a subtle pattern of cognitive changes over early childhood
may predict schizophrenia compared to depression and no illness (Brekke et al.
2007). In this study, children aged 7–13 who developed adult schizophrenia
exhibited cognitive impairments that emerged early and remain stable on tests of
verbal and visual knowledge acquisition, reasoning, and conceptualization. They
also demonstrated developmental cognitive growth that was slower relative to
healthy comparison subjects on tests indexing processing speed, attention,
visual–spatial problem solving ability, and working memory. These two premorbid
cognitive patterns were not observed in children who later developed recurrent
depression. The authors concluded that the origins of schizophrenia include two
interrelated developmental processes evident from childhood to early adolescence.
Children who will grow up to develop adult schizophrenia enter primary school
Cognitive Impairment in Schizophrenia 19
struggling with verbal reasoning and lag further behind their peers in working
memory, attention, and processing speed as they get older.
Prospective studies have suggested that cognitive impairment is manifest in
individuals who are identified as being at “ultra-high” risk (Caspi et al. 2003) for
schizophrenia by virtue of their family history of schizophrenia and/or the manifes-
tation of mild signs and symptoms consistent with the prodromal symptoms of
schizophrenia (Yung and McGorry 1996; Brewer et al. 2003). Some aspects of
cognitive and perceptual performance in ultra-high risk individuals have been
found to predict which individuals will develop psychotic symptoms such as
olfactory impairment (Yung and McGorry 1996), verbal memory impairment
(Hawkins et al. 2004), and attentional impairment (Brewer et al. 2005). Data
combined from the seven sites of the North American Prodromal Longitudinal
Study (NAPLS) consortium indicate that poorer scores on an overall composite
score of several tests provided the most sensitive measure that differentiated those
high-risk children who would develop psychosis from those who would not, and
worse verbal memory scores predicted a briefer time to psychosis in those who
developed schizophrenia (Reichenberg et al. 2010). However, when regression
models were used, a clinical cluster of genetic risk for schizophrenia with recent
deterioration in functioning, higher levels of unusual thought content, higher levels
of suspicion/paranoia, greater social impairment, and a history of substance abuse
predicted psychosis best (Keefe et al. 2006b) and cognitive measures did not
contribute additionally beyond the clinical measures.
One of the important limitations of the work completed to date has been a
reliance upon the assessment of cognition in schizophrenia and at-risk states with
measures designed to measure intelligence or brain damage that may not be
sensitive to the specific neural circuitry impairments underlying schizophrenia.
Methodologies investigating the specific cognitive and neurobiological processes
that may underlie and possibly precede the conversion to psychosis are likely to
yield greater risk prediction specificity. Human perception, thought, and action—
the basic elements of maintaining reality—are based upon a hierarchical process
that conjoins memory and external stimuli, which has been referred to as learning-
dependent predictive perception (Cannon et al. 2008; Keefe et al. 2011b). It has
been hypothesized that perturbations of the circuitry underlying learning-dependent
predictive perception may contribute to risk for developing schizophrenia and thus
early detection of risk may be more successful with tasks specifically designed to
test memory-prediction function (Krishnan et al. 2011a; Keefe et al. 2011c; Kraus
et al. 2009).
1.7 Assessment of Cognition in Schizophrenia Treatment Studies
As listed in Table 3, multisite trials present a large number of challenges that need
to be met for cognitive data to be collected reliably and efficiently (Keefe and Kraus
2009). Sites and testers must be trained and certified on the test battery and related
20 R.S.E. Keefe and P.D. Harvey
procedures. Data review processes must be established, followed, and maintained
throughout the course of the study. Plans must be in place for adding replacement
testers or new sites during the study. Test selection must address the scientific
hypotheses of the investigators yet be efficient to implement without excessive
missing data. Finally, the data analytic plan should focus on a single or small number
of outcome measures to reduce statistical errors and avoid reduced statistical power.
1.7.1 Registration (Phase III) Trials
The primary product of the MATRICS project was a battery of tests that could be
used across treatment studies. This battery, the MATRICS Consensus Cognitive
Battery (MCCB), was vetted by a panel of experts in schizophrenia, cognition and
clinical trials, validated (Nuechterlein et al. 2008), and normed for ease of use
(Kern et al. 2008). This battery of tests was chosen on the basis that these tests were
in the key domains of cognition in schizophrenia, had excellent psychometric
properties, relations to functional outcomes, were practical for use in clinical trials,
and were not burdensome for patients (Nuechterlein et al. 2008). The battery
includes ten tests of cognition in seven domains (see Table 1). It was accepted by
the US Food and Drug Agency (FDA) as the primary endpoint for registration trials
of cognition in schizophrenia (Sevy and Davidson 1995; Buchanan et al. 2005). In
multisite government and industry clinical trials, the MCCB has demonstrated
sensitivity to cognitive deficits in all domains, excellent test–retest reliability,
small practice effects, and is strongly correlated with measures of functional
capacity (Buchanan et al. 2011a, b; Keefe et al. 2011a). See Fig. 1. To date,
translations have been made available in over 15 languages.
1.8 Early Phase Trials
While the MCCB has been established as the gold standard for schizophrenia
registration trials, it is possible that earlier phase work may benefit from the use
(5–20 mg/day), quetiapine (200–750 mg/day), or ziprasidone (40–160 mg/day).
This trial produced similar results with no differences between treatments, even in
antipsychotic-naıve patients (Sweet et al. 2000). However, all groups showed a
modest improvement. These improvements were only slightly stronger than prac-
tice effects, and demonstrated a relation to clinical symptom change, suggesting
that first-episode patients may demonstrate some overall cognitive benefit related to
overall clinical improvement. However, recent studies of completely antipsychotic-
naıve patients suggest that while standard neuropsychological measures may dem-
onstrate little change with treatment, other more specific measures of cognitive
neuroscience processes such as speeded saccadic latencies to visual targets are
normalized by risperidone but not haloperidol treatment (Reilly et al. 2006).
Follow-up studies utilizing cognitive neuroscience tasks across specific cognitive
domains may yield useful insights as was observed with the CATIE trials.
Overall, these data suggest that in current treatment settings, the impact of
antipsychotic medications on neurocognition varies little on average, with minimal
benefit for most treatments. The nature of these trials cannot exclude the possibility
that some individual patients experience benefits while others worsen, possibly
differentially across medications, but do suggest that there is no specific medication
to which a switch would ensure benefit.
2.2 Pharmacological Augmentation as a Cognitive EnhancementStrategy
Pharmacological augmentation as a treatment strategy is consistent with best
practices for the treatment of other illnesses. For instance, the treatment of hyper-
tension and heart disease typically involves multiple medications with different
targets, such as diuretics, ACE inhibitors, and calcium channel blockers. In schizo-
phrenia, the analogous treatment might include atypical antipsychotic medications,
treatments for negative symptoms, and treatments for cognitive deficits. Based on
the history of FDA evaluation of treatments for cognitive and functional deficits in
dementia, a model strategy for the development of cognitive enhancing treatments
for schizophrenia has been advanced and endorsed. As a result of the MATRICS
initiative, a unique collaboration between the FDA, the National Institute of Mental
Health (NIMH), academia, and the pharmaceutical industry and a consensus
regarding the acceptable methods for conducting a registration trial were developed
(Sevy and Davidson 1995) and modified (Buchanan et al. 2005). There are several
critical features of this design.
2.2.1 Clinical Stability
The FDA has long been concerned that new treatments that improve cognition do so
directly, rather than by reducing the severity of other features of the illness. Thus, a
Cognitive Impairment in Schizophrenia 27
treatment that improves cognition must do so in the absence of improvements in
other illness features, such as psychosis. Since FDA has thus far taken the position
that simultaneous changes in illness features (cognition and psychosis) that are not
statistically correlated may be related, only patients who are clinically stable can
participate. This screening criterion was initially defined as a moderate or less (<4)
severity rating on selected PANSS positive scale items at both screening and
baseline (Sevy and Davidson 1995), but has recently been revised to allow patients
who receive a score of 5 on the PANSS positive items (Buchanan et al. 2005). Also,
there can be no hospitalization for psychiatric illness for at least 8 weeks prior to
screening.
2.2.2 Treatment Stability
This is defined by no major change in antipsychotic medications for at least 6 weeks
prior to screening.
2.2.3 No Medications That Can Influence Cognitive Functioning
This is defined by no treatment with anticholinergics, amphetamines, or L-DOPA.
2.2.4 Treatment Duration
At least some of the pivotal trials must have a 6-month treatment duration. This
requirement is based on the idea that treatment effects must be durable and is
influenced by concerns that the benefit of certain treatments may not persist over
time. However some evidence indicates that cognitive enhancing treatments in
people with schizophrenia can have benefits that occur within minutes to hours
(Carter and Barch 2007).
2.2.5 Co-primary Measure
The FDA required a “co-primary” in cognitive enhancement studies in dementia.
This requirement was designed to ensure that changes in cognition on a
performance-based test led to a clinically meaningful change in everyday function-
ing. In the context of, for instance, cholinesterase inhibitor treatment of dementia,
this requirement makes sense because none of the treatments approved by the FDA
actually led to immediate improvements in functioning, but rather treatments were
deemed successful for suspending the otherwise inexorable decline seen in
Alzheimer’s disease.
Similarly, a co-primary measure has been required for schizophrenia cognitive
enhancement trials. However, there is little evidence that any of the currently
28 R.S.E. Keefe and P.D. Harvey
available co-primary measures have the potential to be sensitive to treatment-
related changes in performance. The existence of this FDA requirement led to a
comprehensive collaborative study, funded by grants from the pharmaceutical
industry to the Foundation for the National Institute of Mental Health (F-NIMH),
which was recently completed, presented to the public, and is now published. The
results of that study (Green et al. 2011) indicated that performance-based measures
of functional capacity were clearly superior to interview-based assessments of
cognitive functioning in terms of their convergence with the MCCB. It needs to
be stressed that this was a cross-sectional validation study and not a treatment
outcomes study.
2.3 Results of Cognitive Enhancement Efforts to Date
Several cognitive enhancement treatment research programs with a wide variety of
treatment mechanisms are under way. Very recent data from Phase II trials suggest
that some compounds may have promise for improving cognition in schizophrenia,
but none of these compounds have been approved for actual use in patients. Some
of these studies have been completed with negative results (Keefe et al. 2011c).
While a full discussion of the reasons for the negative results would be speculative
and premature, one of the major issues that may be important is that of possible
interfering effects of antipsychotic medications. A single abnormal neurotransmit-
ter system is unlikely to lead to the widespread impairments seen, but it is quite
likely that single-transmitter interventions could be interfered with by the block-
ading effects of antipsychotic medications. Most importantly, however, many of the
studies completed to date have been seriously underpowered to detect true treat-
ment effects. A recent review of all studies completed as of June 1, 2011 (Keefe
et al. 2011c) suggested that none of the studies above had sufficient power to detect
a medium (d ¼ 0.5) effect size, which would require 71 subjects per group assum-
ing the primary outcome measure has excellent test–retest reliability (ICC ¼�0.90) as with the MCCB composite score (Keefe et al. 2011a). Several studies
had sufficient power to detect a large (d ¼ 0.8) effect.
2.4 Cognitive Remediation as a Platformfor Pharmacologic Studies
While broad-ranging initiatives are ongoing to refine our understanding of the
mechanisms of cognitive improvement in schizophrenia, an additional area of
consideration is the relatively impoverished cognitive lives of patients who enroll
in pharmacologic enhancement studies. It is possible that many of these experi-
mental pharmacologic interventions will be of only minimal benefit when patients
are evaluated in the context of their habitual low level of cognitive stimulation.
Cognitive Impairment in Schizophrenia 29
Part of the explanation for why clinical trials testing the efficacy of cognitive-
enhancing medications have so far been largely unsuccessful may be that patients in
these trials are not provided with substantive opportunity to utilize the cognitive
benefit that they may have acquired during the drug treatment study. Thus, analo-
gous to the need for physical exercise in an individual who takes steroids to increase
muscle mass, schizophrenia patients in pharmacological intervention trials may
require systematic cognitive training to “exercise” any newfound cognitive poten-
tial that they may have acquired from drug treatment (Keefe et al. 2011d).
Cognitive remediation may provide an excellent platform for enriching the cogni-
tive environment of patients engaged in pharmacologic trials to improve cognition.
Several studies and meta-analysis suggest that cognitive remediation produces
medium effect size improvements in cognitive performance and, when combined
with psychiatric rehabilitation, also improves functional outcomes (McGurk et al.
2007a, b). Additionally, patients find these programs to be enjoyable and engaging,
and they have been linked with increases in participant self-esteem (Wykes et al.
1999). Ongoing treatment with cognitive remediationmay thus provide schizophrenia
patients with the necessary cognitive enrichment and motivation to demonstrate the
true potential of effective cognitive enhancement with pharmacologic intervention.
Recent work suggests that these methods are feasible in clinical trials even at sites
without cognitive remediation experience (Keefe et al. 2012).
3 Conclusions
Cognitive functioning is moderately to severely impaired in patients with schizo-
phrenia. This impairment is the prime driver of the significant disabilities in
occupational, social, and economic functioning in patients with schizophrenia.
The profile of deficits in schizophrenia includes many of the most important aspects
of human cognition: attention, memory, reasoning, and processing speed. While
various efforts are under way to identify specific aspects of neurocognition that may
lie closest to the neurobiological etiology and pathophysiology of the illness, and
may provide relevant convergence with animal models of cognition, standard
neuropsychological measures continue to demonstrate the greatest sensitivity to
functionally relevant cognitive impairment. These measures have been the primary
outcome measures in treatment studies, as exemplified by the MCCB.
There have been several prominent negative treatment trials, including large-
scale studies examining the effects of antipsychotic medications on cognition in
schizophrenia and first-episode psychosis. There have also been a number of
prominent negative studies of add-on treatments, although very few of these studies
have had sufficient statistical power to generate firm conclusions. In addition, a few
recent studies examining novel add-on treatments have produced some encouraging
findings. Ongoing work aims to produce more specific cognitive neuroscience
measures that may be more sensitive targets for pharmacologic intervention.
30 R.S.E. Keefe and P.D. Harvey
Cognitive remediation programs have generated considerable interest as these
methods are far less costly than pharmacologic treatment and are likely to be safer.
A growing consensus suggests that these interventions produce modest gains
for patients with schizophrenia, but the efficacy of the various methods used
has not been empirically investigated. An additional consideration for cognitive
remediation methods is that they may serve as an excellent platform of cognitive
enrichment in trials of pharmacologic treatment to generate the cognitive activity
that may be necessary to register pharmacologic benefit.
References
Addington J, Addington D (1998) Facial affect recognition and information processing in schizo-
phrenia and bipolar disorder. Schizophr Res 32:171–181
Addington J, Addington D (2000) Neurocognitive and social functioning in schizophrenia:
a 2.5 year follow-up study. Schizophr Res 44:47–56
Addington J, Addington D, Maticka-Tyndale E (1991) Cognitive functioning and positive and
negative symptoms in schizophrenia. Schizophr Res 5:123–134
Aleman A, Hijman R, de Haan EH, Kahn RS (1999) Memory impairment in schizophrenia:
a meta-analysis. Am J Psychiatry 156:1358–1366
American Psychiatric Association (2000) Diagnostic and statistical manual of mental disorders
DSM-IV-TR, Text Revision 4th edn. American Psychiatric Association, Washington, DC,
ISBN 0-89042-024-6
Baddeley A (1992) Working memory. Science 255:556–559, Review
Barch DM, Keefe RSE (2010) Anticipating DSM-V: Opportunities and challenges for cognition
and psychosis. Schizophr Bull 36:43–47
Bell MD, Bryson G (2001) Work rehabilitation in schizophrenia: does cognitive impairment limit
improvement? Schizophr Bull 27:269–279
Bell MD, Greig TC, Kaplan E, Bryson G (1997) Wisconsin Card Sorting Test dimensions in
schizophrenia: factorial, predictive, and divergent validity. J Clin Exp Neuropsychol
19:933–941
Bilder RM, Mukherjee S, Rieder RO, Pandurangi AK (1985) Symptomatic and neuropsychological
components of defect states. Schizophr Bull 11:409–419
Brekke JS, Long JD, Nesbitt N, Sobel E (1997) The impact of service characteristics on functional
outcomes from community support programs for persons with schizophrenia: a growth curve
analysis. J Consult Clin Psychol 65:464–475
Brekke JS, Kohrt B, Green MF (2001) Neuropsychological functioning as a moderator of the
relationship between psychosocial functioning and the subjective experience of self and life in
schizophrenia. Schizophr Bull 27:697–708
Brekke JS, Hoe M, Long J, Green MF (2007) How neurocognition and social cognition influence
functional change during community-based psychosocial rehabilitation for individuals with
schizophrenia. Schizophr Bull 33:1247–1256
Bressi S, Miele L, Bressi C et al (1996) Deficit of central executive component of working memory
in schizophrenia. New Trends Exp Clin Psychiatry 12:243–252